During the past decade many research groups have reported drug-loaded microbubble formulations. Initially, a hydrophobic anti-tumor agent, paclitaxel, was tested for its ability to associate and retain with the microbubbles shell (Unger et al. 1998; Tartis et al. 2006). Hydrophobic drugs are simply mixed with microbubble shell material and associate with the hydrophobic layer of the shell (Scheme Fig. 12.4a). These drugs are released, possibly in combination with shell fragments, when ultrasound is applied and bubbles are destroyed. Just as traditional ultrasound contrast microbubbles, drug-loaded microbubbles may be visualized with ultrasound imaging, which can provide additional information in the image-guided drug delivery applications.

The endothelial gap junction pores in the normal tissue vasculature are less than 7 nm (Rapoport et al. 2009b), which may limit delivery of nanoparticles and biopharmaceuticals out of the vasculature. In the tumor tissues blood vessels are defective and leaky. In addition, lymphatic drainage in tumor tissue is poor as compared to normal tissues (Matsumura and Maeda 1986). These features allow so-called “passive targeting” to achieve drug and gene delivery and accumulation in the tumors. This phenomenon is known as the enhanced permeability and retention (EPR) effect (Matsumura and Maeda 1986). Microbubbles, with their micrometer dimensions, are optimal as intravascular imaging agents, but their size is too large to penetrate through the vascular endothelial lining in tumors: typically, particles have to be significantly less than 1 μm in diameter to extravasate into the tumor interstitial space (Yuan et al. 1995). Therefore, to deliver drugs to the tissues outside of the vascular bed, nanobubbles (with gas core) and nanodroplets (with liquid core), which are less than 1 μm have been developed. These are generally prepared with sonication. In brief, aqueous solution or nano-dispersion (e.g., micelles) of shell material is sonicated in the presence of perfluorocarbon gas or liquid. Hydrophobic drugs can be loaded in the shell of these particles by hydrophobic interaction. Nanodroplets are usually prepared from liquid perfluoropentane and perfluorohexane and stabilized with a polymer coat or a lipid monolayer shell (Rapoport et al. 2009a, 2011; Mohan and Rapoport 2010). With submicron size, nanodroplets can exit tumor vasculature after intravenous injection and accumulate in the tumor interstitium via EPR mechanism. The boiling points of perfluoropentane and perfluorohexane, which are utilized as liquid core in nanodroplet, are 29 °C and 56 °C, respectively. Due to surface tension, perfluoropentane nanodroplets may remain liquid even at 37 °C, in “superheated” state. However, they can rapidly convert (shift) to gas bubbles in response to ultrasound (Fig. 12.5).

Therefore, these particles are sometimes called phase-shift nanodroplets or nanoemulsion (Marin et al. 2001). This phase shift can be induced by targeted focused ultrasound in–vivo, for example, to convert the nanodroplets that have accumulated in the tumor tissue by EPR effect after intravenous injection. As a result, drug can be released from the particles only at the site of their phase shift by ultrasound exposure. Resulting gas microbubbles can be observed by ultrasound imaging. Therefore, this system might be useful for imaging control of triggered drug delivery.

Gas-core micro- and nanobubbles and liquid-core nanodroplets usually exhibit a relatively short in-vivo lifetime due to core loss (Shiraishi et al. 2011). Lower molecular weight PFCs (usually gases) are better soluble in water and blood; PFC core is lost from the shell and particles become insensitive to ultrasound. Behavior of nanobubbles and nanodroplets under ultrasound exposure (inertial cavitation with rapid collapse) may become uncontrolled. Consequently, damage to cells and tissues might be induced if acoustic pressure is above the mandated ultrasound contrast imaging Mechanical Index (MI) limits. As a “softer” alternative, Zhang et al. (2014) developed an interesting phase shift “solid–liquid-gas” tri-phase transition particle. It is based on a natural “solid–liquid-gas” tri-phase transition medium, L-menthol as an inner core; hollow mesoporous silica is used as the outer shell. The nanoparticle can continuously generate volatile gas in a relatively mild manner, rather than the conventionally abrupt phase transition induced by boiling of superheated PFC liquid. This controllable solid–liquid-gas tri-phase transition of L-menthol can be attributed to the gradual liquid to volatile gas phase-transition at far below its boiling point (212 °C). The generation of L-menthol gas permits extended enhancement of drug delivery following a single bolus injection. In this paper, they reported prolonged blood circulation time (T_1/2 = 130 min) and continuous accumulation in the tumor during 24 h period. These particles have porous structure, so they can be loaded with hydrophobic and hydrophilic drugs at the same time. Synergistic effects of high intensity focused ultrasound therapy and chemotherapy with these particles could be expected.

This loading style is mostly applied for negatively charged molecules such as nucleic acids and some proteins. Microbubbles or nanobubbles with cationic shells (cationic lipids or polymers) are used. This approach has been applied for almost three decades to prepare liposome-nucleic acid complexes for lipofection. Particles are simply mixed with negatively charged molecules (e.g., plasmid DNA, or oligonucleotides) in low or moderate ionic strength media and complexes are formed by electrostatic interaction (Fig. 12.4b). Electrostatic adsorption onto the bubble shell has been shown to protect nucleic acid from degradation by nucleases, as well as enhance loading efficiency. Net positive charge of the complexes also helps attach the bubbles to cell membrane, and further improve delivery efficiency. This design was first reported by (Unger et al. 1997) who prepared positively charged microbubbles as a modification of MRX-115 microbubbles (lipid shell of that parent formulation is based on zwitterionic dipalmitoyl phosphatidylcholine). In the microbubbles intended for gene delivery, a 1:1 mixture of positively charged dipalmitoyl ethylphosphatidylcholine (DPEPC) and zwitterionic dioleoyl phosphatidylethanolamine was utilized. These microbubbles bound plasmid DNA encoding chloramphenicol acetyl transferase, and could enhance transfection of plasmid DNA into cultured cells when 1 MHz continuous wave ultrasound was applied for 5–30 s.

Many research groups have been utilizing electrostatic binding not only for plasmid DNA but also for oligonucleotides such as antisense, small interfering RNA (siRNA), microRNA, mRNA and minicircles. Various compositions of microbubbles and nanobubbles have been developed and tested as gene delivery systems in-vitro and in-vivo. To improve stability of microbubble formulations, use of longer-chain lipids with higher phase transition temperature was suggested (Christiansen et al. 2003). In that study, a mixture of cationic distearoyl trimethylammonium propane (DSTAP) and neutral distearoyl phosphatidylcholine (DSPC) and poly(ethylene glycol) stearate, with longer-chain stearoyl (C₁₈ acyl) groups, was used for bubble preparation. The resulting shell served as an effective stabilizer for micrometer- size perfluorobutane gas core. Refrigerated storage of these microbubbles for many months in sealed vials under perfluorobutane atmosphere was reported. After complexing with plasmid, they have provided efficient transfection in the insonated skeletal muscle or myocardium in-vivo (Christiansen et al. 2003). Stability of microbubble formulations may be a major factor for gene delivery efficiency (Alter et al. 2009).

The type of cationic molecules is important for efficient transfection, as is the ratio between cationic lipids and nucleic acids. The latter is usually described as N/P ratio. N represents total number of aminogroups (nitrogen atom, positive charge) derived from lipid in bubbles; P represents total number of phosphate groups (phosphorus atom, negative charge) derived from nucleic acids. Optimization of N/P ratio is as important for sonoporation as it is for conventional gene delivery methods based on cationic molecules such as lipofection or polyplex. Takahashi et al. reported that distearoyl dimetylammonium propane (DSDAP)-based cationic liposomal nanobubbles had the highest gene delivery efficiency in comparison with other cationic lipid formulations, such as DSTAP or dimethyldioctadecylammonium bromide (DDAB) (Endo-Takahashi et al. 2013). DSDAP-based plasmid DNA-loaded nanobubbles achieved therapeutic effect in mice with hindlimb ischemia by bFGF gene transfer with ultrasound. Lately, nucleic acid delivery is focused on siRNA and microRNA, the novel gene therapy technologies. For gene silencing, siRNA molecules should be delivered into the most of target cells to inhibit expression of a desired gene. However, a simple co-injection of microbubbles and free nucleic acids results in very low transfection efficiency. Yet even sonoporation with nucleic acid-loaded bubbles would not be sufficient to achieve the desired transfection levels. To overcome this problem, Un et al. (2010a, b, 2011) developed active targeting of nucleic acid-loaded nanobubbles, mannose-modified cationic liposomal bubbles, which utilized mannose-modified PEG-lipids for macrophage targeting. In their initial reports, plasmid DNA-loaded mannose-modified liposomal bubbles provided higher transfection efficiency than non-targeted liposomal bubble (Un et al. 2010a, b). Similar mannosylated sonosensitive siRNA lipoplexes targeted to hepatic endothelium suppressed intracellular adhesion molecule-1 (ICAM-1) expression (Un et al. 2012). Moreover, this group has reported lipopolyplex bubbles, prepared as electrostatic complexes of plasmid DNA and cationic polymer with anionic liposomal nanobubbles (Kurosaki et al. 2014). These bubble lipopolyplexes could reduce cytotoxicity, aggregation with erythrocytes and inflammatory response, the usual problems of cationic gene delivery tools. There is another elegant gene delivery system. Chen et al. developed beta cell-specific gene expression system in pancreatic islets by the combination of ultrasound and plasmid DNA-loaded microbubbles (Chen et al. 2006; Chai et al. 2009). In this study, rat insulin promoter-driven human insulin plasmid DNA and lipofectamine 2000 (commercially available cationic liposomes for gene delivery) were mixed, and the electrostatic complexes were added to phospholipids suspension in the perfluoropropane gas-filled sealed vial and shaken in an amalgamator to prepare microbubbles. Following intravenous infusion of these microbubble-lipofectamine-plasmid complexes, ultrasound was directed at the pancreas to destroy microbubbles within the pancreatic microcirculation, and islet-specific transgene expression in pancreas was observed. Expression peaked at day 4 and then decayed steadily over 4 weeks following a single treatment. Later, these researchers have expanded their studies to sonoporation gene delivery regeneration therapy (Chen et al. 2007, 2012, 2014). Overall, electrostatic drug-microbubble complexes play an important role in the development of nucleic acid sonoporation delivery systems.

There are other drug loading methods, based on covalent coupling between the drug and the bubble. Covalent coupling of biomolecules to microbubbles has been assessed and optimized for targeted molecular ultrasound imaging for almost two decades, as described and reviewed elsewhere (Villanueva et al. 1998; Klibanov 2005). For therapeutic protein delivery applications (Fig. 12.6), Bioley et al. developed antigen-decorated microbubble formulations (Bioley et al. 2012a, b, 2013), and demonstrated that microbubbles can serve as an efficient antigen delivery system to promote phagocytosis of the model antigen ovalbumin (OVA) even without ultrasound exposure: in-vivo administration of OVA-loaded microbubbles in mice resulted in the induction of OVA specific antibody and T cell responses.

There are specialized loading methods to protect nucleic acid from nuclease and improve per-particle loading efficiency. Positively charged liposomes, lipoplexes and polyplexes that entrapped and/or stabilize nucleic acids or drug-loaded nanoparticles, can be placed on the surface of the microbubbles, either via streptavidin-biotin bonds (Vandenbroucke et al. 2008; Yang et al. 2013), or by covalent coupling (Sirsi et al. 2012). In addition, there are reports that describe virus-coated microbubbles. Porter, Bamber et al. (Taylor et al. 2007) developed a microbubble complex with an entry-deficient retrovirus, where positively charged microbubbles carried viral particles attached electrostatically. As mentioned above, these particle-loaded microbubbles were formed by combining pre-formulated microbubbles with particles. Alternative approach is also possible: Bekeredjian et al. prepared adeno-associated virus (AAV) vector-loaded microbubbles by co-amalgamation of viral particles with lipids in water-glycerol solution, by a rapid (45 s) mechanical activation with a Capmix/Vialmix amalgamator and succeeded to develop an ultrasound gene delivery system in a cardiac setting (Muller et al. 2008). Lentacker and colleagues (Geers et al. 2011) developed an in-situ microbubble-liposome complex preparation method, where drug-loaded liposomes (Doxil) were added to glycerol:propyleneglycol:water media containing DSPE-PEG-maleimide, dipalmitoyl phosphatidylcholine (DPPC) and distearoyl-sn-glycero-3-phosphoethanolamine-N-[PDP (polyethylene glycol)-2000] (DSPE-PEG-PDP) in sealed vials with decafluorobutane headspace, followed by the same amalgamation step. This mechanical activation gives rise to C₄F₁₀ lipid-shelled microbubbles decorated with liposomes. The liposomes are attached to the microbubble shell via covalent thiol-maleimide linkages; respective DSPE lipid anchors on the termini of the conjugate are embedded in the liposome bilayer and bubble monolayer shell. Finally, it was reported that doxorubicin-liposome-microbubbles allowed ultrasound-triggered killing of melanoma cells in-vitro even at very low doses of doxorubicin (Escoffre et al. 2013). The advantage of this approach is in the small size, low price and availability of amalgamators. Ease of use and short time of the preparation ensure ability to generate sterile microbubble-particle complex at the bedside.

Liposome is an intelligent drug/gene delivery tool because of its biocompatibility, flexible composition and ease of loading of various types of drugs (including nucleic acids) inside and outside of the particle. Targeting molecules can be placed on the liposome surface for selective drug delivery to receptor-carrying cells and surfaces. Recently, liposome-based nanobubbles were developed. Suzuki, Maruyama and Negishi et al. reported a sonosensitive liposome particle encapsulating a perfluoropropane nanobubble inside it (Fig. 12.7). This liposomal bubble was termed “Bubble liposome” (Suzuki et al. 2007, 2008; Kodama et al. 2010). For plasmid DNA, siRNA and miRNA delivery, they utilized cationic Bubble liposomes loaded with these nucleic acids. As an alternative siRNA-loading method, they used hydrophobic anchor, cholesterol modified siRNA (Negishi et al. 2011). Targeting of Bubble liposomes with specific ligands further enhanced gene delivery efficiency.

A similar formulation with a similar structure, named “eLiposome” was developed recently by Pitt et al. (Fig. 12.8). This particle entrapped perfluoropentane and/or perfluorohexane nanodroplet inside a liposome (Javadi et al. 2012; Lattin et al. 2012a). First, nanoemulsion of perfluoropentane and/or perfluorohexane was prepared. Then nanodroplets were encapsulated in liposomes by co-sonication for the mixture of nanodroplets and liposomes, or by hydration of dry lipid film with the aqueous dispersion of nanodroplets. Finally, nanodroplet encapsulated liposome was purified with step-wise density gradient centrifugation (Javadi et al. 2012). This eLiposome particles were sensitive to ultrasound exposure: co-entrapped calcein release was demonstrated (Lattin et al. 2012b). It is suggested that ultrasound helps convert liquid fluorocarbon to gas, leading to a drastic increase of volume and rupture of the surrounding membrane. Targeted, folate modified eLiposome was developed and demonstrated targeted enhanced delivery of calcein and plasmid DNA into HeLa cells with ultrasound (Javadi et al. 2013). Likewise, folate-modified doxorubicin encapsulated eLiposome was also developed; it could effectively deliver doxorubicin into HeLa cells and provided cytotoxicity enhancement (Lin et al. 2014).